AI Agents Are Starting To Rewrite The Software Industry

Enterprise spending on AI-native software is now growing far faster than traditional cloud software, signalling a major change in how businesses buy, use, and value technology.

Why The Traditional SaaS Model Is Under Pressure

For more than two decades, most enterprise software has operated on a relatively simple model. Businesses bought software licences based on the number of employees using a platform, often referred to as “per-seat” pricing.

This approach helped drive the growth of companies such as Salesforce, Workday, ServiceNow, Slack, Zoom, and countless other Software-as-a-Service (SaaS) providers. Revenue grew as customers added more staff and purchased more licences.

However, the rapid rise of AI agents and AI-native platforms is starting to disrupt that model.

Instead of simply giving employees tools to work with, AI-native systems increasingly aim to complete tasks themselves. For example, AI agents can now respond to customer enquiries, generate marketing campaigns, summarise meetings, analyse contracts, process onboarding requests, monitor systems, and automate internal workflows with limited human involvement.

This changes the economics of enterprise software because companies may no longer need as many human users interacting directly with traditional platforms.

The Spending Gap Is Growing Quickly

One clear sign of this transition comes from procurement platform Tropic, which analysed more than $18 billion in managed software spending. Its latest figures show AI-native enterprise spending grew by approximately 94 per cent year-on-year among mid-market and enterprise organisations, while primarily traditional SaaS spending grew by around eight per cent.

It’s important to note that these figures reflect Tropic’s customer dataset rather than the entire global software market. However, analysts increasingly believe the underlying trend is real and accelerating.

Also, research from Deloitte suggests software companies are now under growing pressure to become “AI-first” businesses, with agentic AI expected to transform software operations, pricing models, and customer expectations across the industry.

Meanwhile, Gartner predicts that by 2030, at least 40 per cent of enterprise SaaS spending could move towards usage-based, agent-based, or outcome-based pricing models instead of traditional per-seat licensing.

What “AI-Native” Actually Means

Much of the current discussion centres around the difference between traditional SaaS, hybrid AI software, and fully AI-native systems.

Traditional SaaS platforms mainly rely on human users manually operating software interfaces. Hybrid systems add AI features into existing platforms, such as AI assistants inside Microsoft 365 or Salesforce.

AI-native platforms are different because the AI itself becomes the main worker inside the system.

For example, some newer customer service platforms now allow businesses to deploy autonomous AI agents capable of handling large volumes of enquiries across WhatsApp, email, web chat, and social media with minimal human input. Other AI-native systems can build workflows, generate reports, write software code, or analyse data through natural language instructions rather than manual configuration.

This helps explain why investors and software vendors are increasingly focusing on “agentic AI”, where software performs work autonomously rather than simply assisting humans.

Why Software Companies Are Rushing To Adapt

The pressure on traditional software firms is now becoming increasingly visible.

Many major software providers are rapidly embedding AI agents into their products, partly because investors fear that platforms failing to adopt AI quickly enough could lose market share to newer AI-native competitors.

Salesforce, Microsoft, Google, ServiceNow, Slack, Anthropic, OpenAI, and many others are now heavily promoting AI agents and autonomous workflow systems as core parts of their future strategies.

If one AI agent can perform work that previously required several employees using multiple software licences, the traditional per-user revenue model that has underpinned much of the software industry for decades becomes harder to sustain.

This has also created growing interest in alternative pricing structures based on usage, AI actions, outcomes, or completed tasks rather than simply employee headcount.

At the same time, many businesses are discovering that AI systems introduce very different cost structures from traditional SaaS.

Unlike standard software subscriptions, AI systems often consume large amounts of compute power, tokens, API calls, and cloud infrastructure. Research cited by Tropic suggests many organisations are now seeing AI-related software price increases far above normal annual SaaS uplifts.

What Does This Mean For Your Business?

For UK businesses, the most important point is that AI is increasingly moving beyond being a standalone productivity tool and is starting to reshape the software industry itself.

Businesses evaluating software suppliers may increasingly need to ask not just what a platform does, but how much human work it can realistically automate, what the long-term pricing model looks like, and how AI-generated decisions are monitored and controlled.

The trend also means software procurement is becoming more complicated. Traditional, predictable per-user pricing is gradually being replaced by models based on AI usage, actions, compute consumption, or business outcomes, which may make long-term costs harder to forecast.

At the same time, organisations adopting AI-native systems may gain significant efficiency advantages if these tools genuinely reduce manual workload, improve customer response times, or automate repetitive operational tasks.

However, many AI agents still remain imperfect, requiring human oversight, careful governance, and strong security controls. Businesses should therefore be cautious about assuming that AI-native automatically means lower risk or lower cost.

What is becoming increasingly clear, however, is that the software industry is entering a major transition period. The companies that succeed may not necessarily be those with the biggest software platforms, but those that can most effectively combine AI automation, workflow integration, trust, and measurable business outcomes into products organisations are willing to rely on every day.